X-pop3-spooler:	POP3MAIL 2.1.0 b 4 980420 -bs-
Delivered-To:	pcg AT goof DOT com
Date:	Mon, 27 Apr 1998 00:00:13 +0200 (MET DST)
From:	Fabio Coatti <cova AT felix DOT unife DOT it>
X-Sender:	cova AT cova DOT fe DOT lix
Reply-To:	Fabio Coatti <cova AT felix DOT unife DOT it>
To:	beastium-list AT Desk DOT nl
Subject:	tests
Message-ID:	<Pine.LNX.3.96.980426230100.25071A-100000@cova.fe.lix>
MIME-Version:	1.0
Sender:	Marc Lehmann <pcg AT goof DOT com>
Status:	RO
Lines:	84

Last week I was prone to do some experiments, so I've installed
pgcc-1.0.2.linux86.tar.bz2 and pg++-1.0.2.linux86.tar.bz2 bianries on my linux
libc5-based system (P133Intel), then I've compiled and installed the 
glibc-2.0.6 package (also threads, crypt and locale) as primary libs. I've compiled the
glibc with -O6 -mpentium otpions, and all seems to work fine. After this,
I've downloaded egcs-1.0.2 and pgcc1.0.2 diff (pgcc2.90.27), and compiled with make
bootstrap to obtain a glibc2-based compiler. I've also forced -O6 -mpentium
options. I've some problems with configure (if I usee the --prefix option
to set a path different from /usr/local, I'm unable to compile the compiler,
so now I've my gcc compiler installed under /usr/local/bin), but I've some
FAILS in make check. Here the logs report:

XFAIL: gcc.c-torture/compile/961203-1.c
FAIL: gcc.c-torture/execute/strct-pack-3.c execution,  -O2 
-fomit-frame-pointer -finline-functions -funroll-loops
FAIL: gcc.c-torture/execute/strct-pack-3.c execution,  -O2
-fomit-frame-pointer -finline-functions -funroll-all-loops
XFAIL: gcc.failure/940409-1.c,  -O0
XFAIL: gcc.failure/940409-1.c,  -O1
XFAIL: gcc.failure/940409-1.c,  -O2
XFAIL: gcc.failure/940409-1.c,  -O2 -fomit-frame-pointer -finline-functions

                === gcc Summary ===

# of expected passes            4881
# of unexpected failures        2
# of expected failures          5
# of unsupported tests          7
/d2/usr/src/egcs-1.0.2/objdir/gcc/xgcc version pgcc-2.90.27 980315
(egcs-1.0.2 release)

And for g++:
XFAIL: g++.benjamin/warn01.C (test for excess errors)
XFAIL: g++.brendan/bit-fields2.C .* , (test for errors, line 11)
XFAIL: g++.bugs/900213_03.C missed by g++, warned by cfront, (test for errors, line 24)
XFAIL: g++.bugs/900215_02.C , (test for errors, line 44)
XFAIL: g++.bugs/900321_01.C , (test for errors, line 22)
XFAIL: g++.bugs/900321_01.C , (test for errors, line 23)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 25)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 29)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 35)
[...some other XFAIL]
FAIL: g++.mike/p6610a.C  Execution test


                === g++ Summary ===


# of expected passes            3399
# of unexpected failures        1
# of unexpected successes       1
# of expected failures          82
# of untested testcases         6
/d2/usr/src/egcs-1.0.2/objdir/gcc/testsuite/../xgcc version pgcc-2.90.27
980315 (egcs-1.0.2 release)

and for g77 a big bunch of failures...
                === g77 Summary ===

# of expected passes            40
# of unexpected failures        50
# of untested testcases         42


It's a known problem or It's my fault? Should I try a lower optimization level
(say -O3, -O2)?

Ignoring this problem, I've installed the libc6 pgcc and compiled some
packages, like bash2.02,gnupg,the whole kde distribution... I've noticed
some problems only in kde, but this soft is in beta stage, so I'm unable to
detect if the problem is in source code or in compilation.
If someone needs some binaries linked with glibc2 and compiled with -O6
-mpentium, I can send it (tar, gzip, grep, bzip2, etc..).


--
-----------------------------------------------------------
Fabio Coatti
2:332/409.414                   Fidonet
cova AT felix DOT unife DOT it             Internet
http://felix.unife.it/~cova     Home page
Old SysOps never die... they simply forget their password.
-----------------------------------------------------------

- Raw text -